Post by thread:[EE] voltage monitors

I've got a solar charging solution that runs a small circuit. It uses a single 1.2v NiMH battery and my goal is to maximize the NiMH cycles. I got one of these slick little voltage supervisory chips: www.mouser.com/Search/ProductDetail.aspx?qs=%2fugrpAKHX8rrqC4E3QUSBg% 3d%3d That triggers at 1.1v - thus shutting off my circuit as soon as the NiMH battery gets to the steep part of the discharge curve. The supervisory drives a 2N2222 to pull the circuit to ground. The problem is that as soon as the device triggers and shuts down the circuit the voltage in the battery climbs back up to 1.2v and the circuit turns back on, pulling the battery down below 1.1 so it turns off, and repeats this probably a dozen times before finally settling below the threshold Is there a simple analog circuit I could wrap around the 2222 to keep it pulled down for several hours even if the supervisory turns on and off? Keep in mind that I don't have a uController available. (although if I could find a very inexpensive 1.2v controller that might be a solution). Thank you, DougM

I'd switch the 2n2222 for a 2n7000. This will eliminate the B-E drain through the 2n2222, and permit the use of a simple r/c network on the gate of the 2n7000 to achieve the delay. Plus, it should be more efficient without the B-E current. I'd probably start with the open drain output version of the supervisory circuit, and connect it from the gate to ground. Put a capacitor in parallel (also from gate to ground). Add a high-value pull up resistor to +V. That way, when it triggers, it will yank the gate to ground, turning off the 2n7000. When it "untriggers", the capacitor will slowly charge through the pullup until it reaches the gate threshold, at which point the 2n7000 will turn back on. -forrest Doug Metzler wrote: {Quote hidden}> I've got a solar charging solution that runs a small circuit. It uses a > single 1.2v NiMH battery and my goal is to maximize the NiMH cycles. > > > > I got one of these slick little voltage supervisory chips: > > > > www.mouser.com/Search/ProductDetail.aspx?qs=%2fugrpAKHX8rrqC4E3QUSBg% > 3d%3d > > > > That triggers at 1.1v - thus shutting off my circuit as soon as the NiMH > battery gets to the steep part of the discharge curve. > > > > The supervisory drives a 2N2222 to pull the circuit to ground. > > > > The problem is that as soon as the device triggers and shuts down the > circuit the voltage in the battery climbs back up to 1.2v and the circuit > turns back on, pulling the battery down below 1.1 so it turns off, and > repeats this probably a dozen times before finally settling below the > threshold > > > > Is there a simple analog circuit I could wrap around the 2222 to keep it > pulled down for several hours even if the supervisory turns on and off? > > > > Keep in mind that I don't have a uController available. (although if I > could find a very inexpensive 1.2v controller that might be a solution). > > > > Thank you, > > > > DougM > > > >

Probably need something more modern with a lower gate threshold than the '7000 since isn't this all running on 1.2 volts? Cheerful regards, Bob Forrest Christian wrote: {Quote hidden}> I'd switch the 2n2222 for a 2n7000. This will eliminate the B-E drain > through the 2n2222, and permit the use of a simple r/c network on the > gate of the 2n7000 to achieve the delay. Plus, it should be more > efficient without the B-E current. > > I'd probably start with the open drain output version of the supervisory > circuit, and connect it from the gate to ground. Put a capacitor in > parallel (also from gate to ground). Add a high-value pull up resistor > to +V. > > That way, when it triggers, it will yank the gate to ground, turning off > the 2n7000. When it "untriggers", the capacitor will slowly charge > through the pullup until it reaches the gate threshold, at which point > the 2n7000 will turn back on. > > -forrest > > Doug Metzler wrote: >> I've got a solar charging solution that runs a small circuit. It uses a >> single 1.2v NiMH battery and my goal is to maximize the NiMH cycles. >> >> >> >> I got one of these slick little voltage supervisory chips: >> >> >> >> www.mouser.com/Search/ProductDetail.aspx?qs=%2fugrpAKHX8rrqC4E3QUSBg% >> 3d%3d >> >> >> >> That triggers at 1.1v - thus shutting off my circuit as soon as the NiMH >> battery gets to the steep part of the discharge curve. >> >> >> >> The supervisory drives a 2N2222 to pull the circuit to ground. >> >> >> >> The problem is that as soon as the device triggers and shuts down the >> circuit the voltage in the battery climbs back up to 1.2v and the circuit >> turns back on, pulling the battery down below 1.1 so it turns off, and >> repeats this probably a dozen times before finally settling below the >> threshold >> >> >> >> Is there a simple analog circuit I could wrap around the 2222 to keep it >> pulled down for several hours even if the supervisory turns on and off? >> >> >> >> Keep in mind that I don't have a uController available. (although if I >> could find a very inexpensive 1.2v controller that might be a solution). >> >> >> >> Thank you, >> >> >> >> DougM >> >> >> >> >

{Quote hidden}> I got one of these slick little voltage supervisory chips: > www.mouser.com/Search/ProductDetail.aspx?qs=%2fugrpAKHX8rrqC4E3QUSBg% > 3d%3d > That triggers at 1.1v - thus shutting off my circuit as > soon as the NiMH > battery gets to the steep part of the discharge curve. > The supervisory drives a 2N2222 to pull the circuit to > ground. > The problem is that as soon as the device triggers and > shuts down the > circuit the voltage in the battery climbs back up to 1.2v > and the circuit > turns back on, pulling the battery down below 1.1 so it > turns off, and > repeats this probably a dozen times before finally > settling below the > threshold > Is there a simple analog circuit I could wrap around the > 2222 to keep it > pulled down for several hours even if the supervisory > turns on and off? Data sheet http://www.onsemi.com/pub_link/Collateral/NCP304-D.PDF One way is to add hysteresis to the sensed voltage. A simple way (that possibly works) is to add a series resistor to the sense IC input / power supply. When operated it's current drain will rise, thereby decreasing the input voltage. Consult data sheet and add an appropriate R. Too large and it won't rewake well when voltage rises. For extra points and more complexity you could have a voltage divider at the input with the transistor adding to the loading. More points from using a time delay (aka capacitor) to make the extra load fully or partially time dependant. Just the series R will probably help. Russell

Yeah, you're right... I think my brain was stuck on the 2n2222 B-E voltage drop for some reason, and not the 2V-ish typical threshold of a 2n7000. I'll claim sleep deprivation as my excuse (only ~4-5 hours last night). That said I'm sure Zetex (and others) have a suitable low-threshold fet which should work for the application as described. -forrest Bob Blick wrote: {Quote hidden}> Probably need something more modern with a lower gate threshold than the > '7000 since isn't this all running on 1.2 volts? > > Cheerful regards, > > Bob > > Forrest Christian wrote: > >> I'd switch the 2n2222 for a 2n7000. This will eliminate the B-E drain >> through the 2n2222, and permit the use of a simple r/c network on the >> gate of the 2n7000 to achieve the delay. Plus, it should be more >> efficient without the B-E current. >> >> I'd probably start with the open drain output version of the supervisory >> circuit, and connect it from the gate to ground. Put a capacitor in >> parallel (also from gate to ground). Add a high-value pull up resistor >> to +V. >> >> That way, when it triggers, it will yank the gate to ground, turning off >> the 2n7000. When it "untriggers", the capacitor will slowly charge >> through the pullup until it reaches the gate threshold, at which point >> the 2n7000 will turn back on. >> >> -forrest >> >> Doug Metzler wrote: >> >>> I've got a solar charging solution that runs a small circuit. It uses a >>> single 1.2v NiMH battery and my goal is to maximize the NiMH cycles. >>> >>> >>> >>> I got one of these slick little voltage supervisory chips: >>> >>> >>> >>> www.mouser.com/Search/ProductDetail.aspx?qs=%2fugrpAKHX8rrqC4E3QUSBg% >>> 3d%3d >>> >>> >>> >>> That triggers at 1.1v - thus shutting off my circuit as soon as the NiMH >>> battery gets to the steep part of the discharge curve. >>> >>> >>> >>> The supervisory drives a 2N2222 to pull the circuit to ground. >>> >>> >>> >>> The problem is that as soon as the device triggers and shuts down the >>> circuit the voltage in the battery climbs back up to 1.2v and the circuit >>> turns back on, pulling the battery down below 1.1 so it turns off, and >>> repeats this probably a dozen times before finally settling below the >>> threshold >>> >>> >>> >>> Is there a simple analog circuit I could wrap around the 2222 to keep it >>> pulled down for several hours even if the supervisory turns on and off? >>> >>> >>> >>> Keep in mind that I don't have a uController available. (although if I >>> could find a very inexpensive 1.2v controller that might be a solution). >>> >>> >>> >>> Thank you, >>> >>> >>> >>> DougM >>> >>> >>> >>> >>>

I looked for a fet with a lower gate voltage and wasn't able to find one, but I'll renew my search. That definitely was the solution I was after because of the transistor drain that you mentioned. Thinking about it the cap and large resistor solution might be a little tricky since with no drain on the battery the voltage is likely to come back up to 1.2 and stay there. Maybe there's a problem with my whole approach - I guess I really need something that will shut off when the supervisory kicks, and not turn on again until it detects a positive flow from the solar panel to the battery. I'll go back to the drawing board and think about this some more - all suggestions are welcome. Thanks! DougM On Wed, Jul 30, 2008 at 10:46 PM, Forrest Christian <spam_OUTforrestcTakeThisOuTimach.com>wrote: {Quote hidden}> Yeah, you're right... I think my brain was stuck on the 2n2222 B-E > voltage drop for some reason, and not the 2V-ish typical threshold of a > 2n7000. I'll claim sleep deprivation as my excuse (only ~4-5 hours > last night). > > That said I'm sure Zetex (and others) have a suitable low-threshold fet > which should work for the application as described. > > -forrest > > Bob Blick wrote: > > Probably need something more modern with a lower gate threshold than the > > '7000 since isn't this all running on 1.2 volts? > > > > Cheerful regards, > > > > Bob > > > > Forrest Christian wrote: > > > >> I'd switch the 2n2222 for a 2n7000. This will eliminate the B-E drain > >> through the 2n2222, and permit the use of a simple r/c network on the > >> gate of the 2n7000 to achieve the delay. Plus, it should be more > >> efficient without the B-E current. > >> > >> I'd probably start with the open drain output version of the supervisory > >> circuit, and connect it from the gate to ground. Put a capacitor in > >> parallel (also from gate to ground). Add a high-value pull up resistor > >> to +V. > >> > >> That way, when it triggers, it will yank the gate to ground, turning off > >> the 2n7000. When it "untriggers", the capacitor will slowly charge > >> through the pullup until it reaches the gate threshold, at which point > >> the 2n7000 will turn back on. > >> > >> -forrest > >> > >> Doug Metzler wrote: > >> > >>> I've got a solar charging solution that runs a small circuit. It uses > a > >>> single 1.2v NiMH battery and my goal is to maximize the NiMH cycles. > >>> > >>> > >>> > >>> I got one of these slick little voltage supervisory chips: > >>> > >>> > >>> > >>> > www.mouser.com/Search/ProductDetail.aspx?qs=%2fugrpAKHX8rrqC4E3QUSBg% > >>> 3d%3d > >>> > >>> > >>> > >>> That triggers at 1.1v - thus shutting off my circuit as soon as the > NiMH > >>> battery gets to the steep part of the discharge curve. > >>> > >>> > >>> > >>> The supervisory drives a 2N2222 to pull the circuit to ground. > >>> > >>> > >>> > >>> The problem is that as soon as the device triggers and shuts down the > >>> circuit the voltage in the battery climbs back up to 1.2v and the > circuit > >>> turns back on, pulling the battery down below 1.1 so it turns off, and > >>> repeats this probably a dozen times before finally settling below the > >>> threshold > >>> > >>> > >>> > >>> Is there a simple analog circuit I could wrap around the 2222 to keep > it > >>> pulled down for several hours even if the supervisory turns on and off? > >>> > >>> > >>> > >>> Keep in mind that I don't have a uController available. (although if I > >>> could find a very inexpensive 1.2v controller that might be a > solution). > >>> > >>> > >>> > >>> Thank you, > >>> > >>> > >>> > >>> DougM > >>> > >>> > >>> > >>> > >>> > > -

>I looked for a fet with a lower gate voltage and wasn't >able to find one, > but I'll renew my search. Digiket etc FDN337 N Channel FDN338 P Channel & Zetex do some. & the lovely CES2310 - but probably Asia sources only. Try a series resistor in the supervisors input though. And, you can add as much load as you want to the output to increase the hysteresis as you wish. MAY need a small cap at superviror input when R used. MADE UP EXAMPLE ONLY Supervisor Iq = 10 uA. Superviros Ion = 100 uA. Transistor load = 100 uA. Using a 1k input resistor drops - 0.01V at 10 uA - 0.2 V with 100 uA IC + 100 uA transistor. If you decided you wanted more drop with output on you could either - Raise Rin to say 3k3 or - Add a load resistor to output, or lower base drive R. Adding input R takes less current from battery when low. Russell

Thank you Russell, I'm going to order some FDN337's and play around with this, but I am going to go back and look at my approach in the context of the whole circuit and see if I can figure out some way to snap it off and not have it turn back on until the solar panel starts generating current. If I can nail this it will help an another larger circuit that I am currently controlling using a 12F683. you guys are great! DougM On Thu, Jul 31, 2008 at 10:42 AM, Apptech <.....apptechKILLspam@spam@paradise.net.nz> wrote: {Quote hidden}> >I looked for a fet with a lower gate voltage and wasn't > >able to find one, > > but I'll renew my search. > > Digiket etc > > FDN337 N Channel > FDN338 P Channel > > & Zetex do some. > > & the lovely CES2310 - but probably Asia sources only. > > Try a series resistor in the supervisors input though. > And, you can add as much load as you want to the output to > increase the hysteresis as you wish. > MAY need a small cap at superviror input when R used. > > MADE UP EXAMPLE ONLY > > Supervisor Iq = 10 uA. > Superviros Ion = 100 uA. > Transistor load = 100 uA. > > Using a 1k input resistor drops > > - 0.01V at 10 uA > - 0.2 V with 100 uA IC + 100 uA transistor. > > If you decided you wanted more drop with output on you could > either > > - Raise Rin to say 3k3 > or > - Add a load resistor to output, or lower base drive R. > > Adding input R takes less current from battery when low. > > > Russell > > > > -

> Thank you Russell, I'm going to order some FDN337's and > play around with > this, > but I am going to go back and look at my approach in the > context of > the whole circuit and see if I can figure out some way to > snap it off and > not have it turn back on until the solar panel starts > generating current. Try the series input resistor that I suggested and see how it meets your needs. That will "snap it off" by a controlled amount. The IC will see a drop in input of (Ioperating - Ioff) / Rseries when the IC turns on. If that is more than the battery "recovery" voltage then it will stay off. When the battery charges again it will turn back on. A method of detecting charging is to place a transistor Vbe junction across the diode in series with the panel. Some extra biasing may be needed. Russell {Original Message removed}

The FDN337N seems to not have the usual internal Zener between Gate and Source for protection. Discovered that hard fact after 12 of them stopped working without any reason. Mark Jordan On 1 Aug 2008 at 5:42, Apptech wrote: {Quote hidden}> >I looked for a fet with a lower gate voltage and wasn't > >able to find one, > > but I'll renew my search. > > Digiket etc > > FDN337 N Channel > FDN338 P Channel > > & Zetex do some. > > & the lovely CES2310 - but probably Asia sources only. > > Try a series resistor in the supervisors input though. > And, you can add as much load as you want to the output to > increase the hysteresis as you wish. > MAY need a small cap at superviror input when R used. > > MADE UP EXAMPLE ONLY > > Supervisor Iq = 10 uA. > Superviros Ion = 100 uA. > Transistor load = 100 uA. > > Using a 1k input resistor drops > > - 0.01V at 10 uA > - 0.2 V with 100 uA IC + 100 uA transistor. > > If you decided you wanted more drop with output on you could > either > > - Raise Rin to say 3k3 > or > - Add a load resistor to output, or lower base drive R. > > Adding input R takes less current from battery when low. > > > Russell > > > > --